Comparison of visual field using standard automated and kinetic perimetry in patients with chiasmal compression. Correlation with peripapillary retinal nerve fiber layer

Abstract

Purpose : To compare visual field (VF) in patients with chiasmal compression with 24-2 standard automated perimetry (SAP) and semiautomated kinetic perimetry (SKP) and assess the relationship between peripapillary retinal nerve fiber layer (pRNFL) measurements and VF loss on SAP and SKP.

Methods : Thirty 30 eyes with temporal VF loss from chiasmal compression (CC) and 30 healthy eyes (CT) underwent OCT optic disc scanning, Humphrey 24-2 SAP and Octopus 900 SKP with V/4e, I/4e, I/3e and I/2e targets. pRNFL thickness measurements were evaluated globally as well as in 4-disc quadrants. VF data from SAP and SKP were evaluated globally and in 4 VF sectors projected to the corresponding superior, inferior, temporal (T), and nasal disc sectors. Groups were compared using Generalized Estimation Equations. We assessed the relationship between functional and structural parameters, using VF parameters as independent variables and pRNFL thickness as the dependent variables. We evaluated the effect of one single VF predictor using single linear regression and using 5 VF predictors using multiple linear regression analysis. Significance was set at p ≤ 5%.

Results : All pRNFL thickness parameters were significantly reduced in CC than CT. SAP and VF area on SKP differed significantly between CC and CT except for the data from VF area that projected to the T disc sector with V4e and I4e. Linear regression evaluating the effect of a single predictor showed significant correlation coefficients when using all SAP parameters (range 0.58 to 0.83) and all (range 0.33 to 0.76) but the central VF x T disc sector with V/4e and I/4e targets. Intraclass correlation coefficients were greater (range 0.60-0.86) and mean error smaller (range 11.0%-20.8%) in the multiple linear regression model using SAP and 4 SKP isopters than in the single linear regression using SAP alone (range 0.55-0.86, 12.4%-23.6%, respectively). Experimental error was smaller in the multiple regression model.

Conclusions : Adding data from both SAP and SKP (isopters V/4e, I4e, I/3e and I/2e) leads to improved correlation and better predicts pRNFL in patients with chiasmal compression. These suggest that adding peripheral VF assessment to SAP may me important to assess the functional-structural relationship in compressive visual pathway diseases.

This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.